1. Field of the Invention
This invention relates to a method of desulfurization of hydrocarbons and particularly to a method of desulfurization of hydrocarbons that uses an efficient, selective oxidation and removal of sulfur- and nitrogen-containing compounds from petroleum distillates wherein the physical properties of the fuel either remain constant or improved.
2. Description of Related Art
Environmental concerns have driven the need to remove many impurities from hydrocarbon based distillate fuels. Sulfur- and nitrogen-containing compounds are of particular interest because of their tendencies to produce precursors to acid rain and airborne particulate material. Several processes have been proposed in the past to deal with the problem of removing of these compounds from fuels. The most prevalent and common industrial process is that of treating the fuel under high temperatures and high pressures with hydrogen. This process is called hydrotreating and has received extensive attention since its original invention in Germany before the Second World War. Literature describing this technology is immense, amounting to thousands of patents and scientific and engineering publications.
Briefly stated, hydrotreating is a process in which a petroleum fraction is heated, mixed with hydrogen, and fed to a reactor packed with a particulate catalyst. Temperatures in the reactor typically range from 600 to 700 F. (315 to 370 C.). At these temperatures, some or all of the feed may vaporize, depending on the boiling range of the feed and the pressure in the unit. For heavier feeds it is common for the majority of the feed to be liquid. Reaction pressures range from as low as 500 psig (pounds per square inch, gauge) to as high as 2500 psig depending on the difficulty of removing the sulfur. In the manufacture of distillate fuels such as diesel or jet fuel, pressures higher than 800 psig are common. The feed and hydrogen mixture typically flows downward through the reactor, passing around and through the particulate catalyst. Upon leaving the reactor, the mixture of treated fuel and hydrogen flows through a series of mechanical devices to separate and recycle the hydrogen, remove poisonous hydrogen sulfide generated in the reaction, and recover the desulfurized product. Hydrotreating catalysts slowly lose activity with use, and must be removed and replaced every two to three years.
As used in large integrated refineries, hydrotreating is very effective and relatively inexpensive. However, in small refineries, and especially those with limited capabilities, it can be prohibitively expensive because of the effects of scale-up economics. When process equipment is built, it typically costs much less than twice as much to build a unit with twice the capacity; engineers typically estimate that doubling the size increases the cost by only about 50%. The converse of the scale-up effect occurs when processes are scaled down; smaller process units are only slightly less expensive to build than larger one. Thus the investment for a small 5,000 barrel per day (bpd) hydrotreater is not 1/10 that of a 50,000 bpd hydrotreater, but is about 1/4 the cost of the much larger unit.
Because of the way processes are operated and controlled, the manpower costs for the smaller unit are roughly the same as those of the larger one.
Another cost problem faced by small refiners is the lack of an inexpensive hydrogen source. Hydrotreating typically consumes 200 to 500 scfb (standard cubic feet per barrel) of hydrogen, and may consume as much as 1000 scfb. Manufacture of hydrogen from natural gas typically costs about $3 per 1000 scf, adding about $0.60 to as much as $3.00 to the cost of treating a barrel of feed for a small refinery. In large refineries, hydrogen is often available as a byproduct of the gasoline manufacturing process known as platinum reforming. As such it is virtually free. In small refineries with no platinum reformer, a dedicated hydrogen manufacturing plant must be installed, adding to the refinery operator's investment burden and operating costs.
These economics favor those who wish to operate at large scale, but they make hydrotreaters prohibitively expensive for smaller refineries. As a result, tightening environmental regulations have had the effect of forcing small refineries to close. Some small refineries have survived by changing product mix to emphasize low value products such as asphalt, selling liquid products to large refineries to use as intermediates.
In order to continue to operate successfully, refineries and others have explored alternatives to hydrotreating. One idea that has been explored involves oxidizing the sulfur and nitrogen compounds in a distillate then removing them by selective extraction. This approach has met with only limited success primarily because of problems of non-selectivity of oxidants or the extraction solvents.
It is known that contacting a distillate with an oxidant, can convert sulfur- and nitrogen-containing compounds to much more polar oxidized species. Such oxidants include peroxy organic acids, catalyzed hydroperoxides, inorganic peroxy acids or peroxy salts. Experience shows that such oxidants are typically those where the predominant oxidation does not include a free radical chain reaction oxidation of the sulfur or nitrogen, but appear to operate by donating oxygen atoms to the sulfur in thiols and thiophenes to form sulfoxides or sulfones, or to the nitrogen in amines, pyridines or pyroles to form nitro, nitroso, or ammine oxide compounds. It is also known that all of these oxidized sulfur- or nitrogen-containing compounds are orders of magnitude more soluble in non-miscible solvents than their unoxidized counterparts.
The next step of this process is removal of the oxidized compounds by contacting the distillate with a selective extraction solvent. This solvent should be sufficiently polar to be selective for polar compounds is the next step of this process. Examples, of polar solvents include those with high values of the Hildebrand solubility parameter .delta.; liquids with a .delta. higher than about 22 have been successfully used to extract these compounds. Examples of polar liquids, with their Hildebrand values, are shown in the following table:
______________________________________ Acetone 19.7 Butyl Cellosolve 20.2 Carbon disulfide 20.5 Pyridine 21.7 Cellosolve 21.9 DMF 24.7 n-Propanol 24.9 Ethanol 26.2 DMSO 26.4 n-Butyl alcohol 28.7 Methanol 29.7 Propylene glycol 30.7 Ethylene glycol 34.9 Glycerol 36.2 Water 48.0 ______________________________________
However, as will be obvious to those skilled in the art, mere polarity considerations are insufficient to define successful extraction solvents. Methanol, for instance, has sufficient polarity, but its density, 0.79 g/cc, is about the same as that of typical hydrocarbon fuels, making separations very difficult. Other properties to consider include boiling point, freezing point, and surface tension. Surprisingly, the combination of properties exhibited by DMSO make it an excellent solvent for extracting oxidized sulfur and nitrogen compounds from liquid fuels.
In U.S. Pat. No. 3,847,800, Guth and Diaz proposed a process for treating diesel fuel that used oxides of nitrogen as the oxidant. However, nitrogen oxides have several disadvantages that can be traced to the mechanism by which they oxidize distillates. In the presence of oxygen, nitrogen oxides initiate a very non-selective form of oxidation termed auto-oxidation. Several side reactions also take place including the creation of nitro-aromatic compounds, oxides of alkanes and arylalkanes, and auto-oxidation products. Oxides of nitrogen are used to synthesize sulfoxides because they tend to inhibit the formation of sulfones due to the presence of oxonium salts. However, for the purposes of sulfur removal from fuels, sulfones are the desired product of sulfur oxidation because of their increased dipole moment, hence, higher solubility in the non-miscible solvent. Thus, nitrogen oxide based oxidants do not yield the appropriately oxidized sulfur compounds in distillate hydrocarbons without creating many undesirable byproducts.
The Guth and Diaz patent also proposes the use of methanol, ethanol, a combination of the two, and mixtures of these and water as an extraction solvent for polar molecules. Although these have proved to be acceptable extraction solvents for this system, they do not perform as well as others.
U.S. Pat. No. 4,746,420, issued to Darian and Sayed-Hamid also proposes the use of a nitrogen oxides to oxidize sulfur- and nitrogen-containing compounds followed by extraction using two solvents--a primary solvent followed by a cosolvent that is different from the primary. The sulfur and nitrogen results published in this patent are consistent with those expected from incomplete oxidation of these compounds followed by extraction.
In European Patent Application number 93302642.9, Method for Recovering Organic Sulfur Compounds from a Liquid Oil, Tetsuo claims many oxidants as being essentially equal in their ability to oxidize sulfur- and nitrogen-containing compounds. However, I have discovered that many of these oxidants are not selective and others are ineffective. Oxidizers that proceed by an auto oxidation mechanism involving a free radical tend not to be selective for the sulfur- and nitrogen-containing compounds of interest, producing numerous side reactions and, hence, various undesirable byproducts.
Tetsuo teaches the use of distillation, solvent extraction, low temperature separation, adsorbent treatment and separation by washing to separate and oxidized organic sulfur compound from the liquid oil through the utilization of differences in the boiling point, melting point and/or solubility between the organic sulfur compound and the oxidized organic sulfur compound. While most of these work with some success, they do not provide the level of sulfur removal that my method achieves.
In "Desulfurization of Petroleum Fractions by Oxidation and Solvent Extraction", Fuel Processing Technology, 42, 1995, 35-45, by F. Zannikos, E. Lois, and S. Stournas, the authors describe an oxidation and solvent extraction technique for the removal of sulfur containing compounds. Peroxyacetic acid was used in an inefficient manner to oxidize the sulfur compounds in a diesel fuel. Methanol, dimethyl formamide, and N-methyl pyrrolidone were used as simple one-stage extraction solvents at different ratios. However, the results of their work show these solvents removed much of the usable oil along with the oxidized sulfur compounds. In order to get sulfur levels of approximately 500 PPM with these solvents they report a loss of 30 or more percent of the overall fuel. Such a loss is completely unacceptable on a commercial basis. No mention of a process is made within this publication. Instead, the authors describe laboratory studies of the oxidation and extraction of sulfur compounds using methods like those taught in the art described above.
Two major problems are seen throughout this art. First, the oxidants chosen do not always perform optimally. Many oxidants engage in unwanted side reactions that reduce the quantity and quality of the treated fuels. The second problem is the selection of a suitable solvent for the extraction of the sulfur or nitrogen compounds. Using the wrong solvent may result in removing desirable compounds from the fuel or extracting less than a desired amount of the sulfur and nitrogen compounds from the fuel. In either case, the results can be costly.